A Noise-Robust Method with Smoothed L1/L 2 Regularization for Sparse Moving-Source Mapping

Mai Quyen I Pham (1) , Benoit I Oudompheng (2, 1) , Jerome I. Mars (1) , Barbara Nicolas (3, 1)

1 GIPSA-SIGMAPHY - GIPSA - Signal Images Physique
2 MicrodB
3 Imagerie Ultrasonore

The method described here performs blind deconvolution of the beamforming output in the frequency domain. To provide accurate blind deconvolution, sparsity priors are introduced with a smoothed 1 // 2 regularization term. As the mean of the noise in the power spectrum domain depends on its variance in the time domain, the proposed method includes a variance estimation step, which allows more robust blind deconvolution. Validation of the method on both simulated and real data, and of its performance, are compared with two well-known methods from the literature: the deconvolution approach for the mapping of acoustic sources, and sound density modeling.

Mots clés: Beamforming blind deconvolution; Acoustic moving-source localization; Smoothed 1 // 2 regularization; Sparse representation; Proximal forward-backward; Robustness algorithms

Domaines: Traitement du signal et de l'image [eess.SP]

Mai Quyen I Pham, Benoit I Oudompheng, Jerome I. Mars, Barbara Nicolas. A Noise-Robust Method with Smoothed L1/L 2 Regularization for Sparse Moving-Source Mapping. Signal Processing, 2016, 135 (June 2017), pp.96-106. ⟨10.1016/j.sigpro.2016.12.022⟩. ⟨hal-01426251⟩